Application of Deep Learning Model in the Sonographic Diagnosis of Uterine Adenomyosis

• DOI: 10.3390/ijerph20031724
• Publication Date: 2023-01-18T06:57:57Z
• Authors: Diego Raimondo; Antonio Raffone; Anna Chiara Aru; Matteo Giorgi; Ilaria Giaquinto; Emanuela Spagnolo; Antonio Travaglino; Federico Andrea G...; Mario Giovanni Co...; Jacopo Lenzi; Gabriele Centini; Lucia Lazzeri; Antonio Mollo; Renato Seracchioli; Paolo Casadio
• Keywords: Adenomyosis

Background: This study aims to evaluate the diagnostic performance of Deep Learning (DL) machine for detection adenomyosis on uterine ultrasonographic images and compare it intermediate ultrasound skilled trainees. Methods: Prospective observational were conducted between 1 30 April 2022. Transvaginal (TVUS) diagnosis was investigated by an experienced sonographer 100 fertile-age patients. Videoclips corpus recorded sequential extracted. Intermediate ultrasound-skilled trainees DL asked make a reviewing images. We evaluated compared accuracy, sensitivity, positive predictive value, F1-score, specificity negative value model diagnosis. Results: Accuracy 0.51 (95% CI, 0.48–0.54) 0.70 0.60–0.79), respectively. Sensitivity, F1-score 0.43 0.38–0.48), 0.82 0.79–0.85) 0.46 (0.42–0.50), respectively, whereas had sensitivity 0.72 0.52–0.86), 0.69 0.58–0.79) 0.55 0.43–0.66). Conclusions: In this preliminary showed lower accuracy but higher in diagnosing intermediate-skilled